A Novel Multi-view Mixture Model Framework for Longitudinal Clustering with Application to ANCA-Associated Vasculitis

2026-04-02

AI Summary

This study addresses the challenge of identifying heterogeneous subgroups in irregularly sampled longitudinal data with respect to disease progression. The authors propose a dual-view probabilistic mixture model that jointly integrates static baseline covariates and dynamic longitudinal biomarker trajectories within a unified clustering framework. Innovatively, the model employs neural ordinary differential equations to capture complex temporal dynamics and incorporates a sparsity-inducing logarithmic penalty into the EM algorithm to enhance subgroup interpretability. Applied to a cohort of ANCA-associated vasculitis patients, the method successfully uncovers distinct patient subgroups characterized by significantly divergent serum creatinine trajectories and markedly different risks of end-stage kidney disease, thereby demonstrating both the methodological validity and clinical utility of the proposed approach.

Abstract

Effectively modeling irregularly sampled longitudinal data is essential for understanding disease progression and improving risk prediction. We propose a two-view mixture model that integrates static baseline covariates and longitudinal biomarker trajectories within a unified probabilistic clustering framework. Temporal patterns are modeled using Neural Ordinary Differential Equations. Model training uses an EM algorithm with a sparsity-inducing log-penalty for interpretable subgroup discovery. Application of the model to an Irish cohort of ANCA-associated vasculitis patients reveals subgroups with heterogeneous serum creatinine trajectories and variation in end-stage kidney disease outcomes.

Problem

• longitudinal clustering
• irregularly sampled data
• disease progression
• heterogeneous subgroups
• ANCA-associated vasculitis

Innovation

• multi-view mixture model
• Neural Ordinary Differential Equations
• longitudinal clustering
• sparsity-inducing penalty
• irregularly sampled data